Using an electronic camera to build a file containing text

• US 6,563,948 B2
• Filed: 04/29/1999
• Issued: 05/13/2003
• Est. Priority Date: 04/29/1999
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Metz, Werner, Tan, Yap-Peng, Acharya, Tinku
• Primary Examiner(s): Patel, Jayanti K.
• Assistant Examiner(s): DESIRE, GREGORY M
• Application Number:
  US09/301,753
• Publication Number:
  US 20020057848A1
• Time in Patent Office: 1,475 Days
• Field of Search: 382/178, 382/298, 382/312-313, 382/321, 382/105, 382/204, 382/291, 340/937
• US Class Current: 382/187
• CPC Class Codes: G06V 30/142 using hand-held instruments; Constructional details of the instruments

Abstract

An embodiment of the invention is directed to a method of building an electronic file, using an electronic camera such as a digital camera, that captures 3-dimensional objects. A number of image data tiles that represent the images are generated by the camera. A number of text data tiles each containing text recognized in a corresponding one of the image data tiles is generated. The method includes searching for overlapping text in the text tiles, and pasting the text tiles in proper alignment into an electronic file.

Claims

1. An electronic camera comprising:

optics having a depth of field greater than a conventional scanner to form a plurality of light images of an object that displays text;

camera imager to receive the light images and in response provide a plurality of image data tiles, respectively, that capture different portions of said text;

optical character recognition engine to provide a plurality of text data tiles each containing a portion of said text recognized from a corresponding one of the plurality of image data tiles; and

text linker to search for a portion of said text that overlaps in the plurality of text tiles; and

means for correcting error text in the plurality of image data tiles, which includes means for re-capturing a region of the scene that contains correct text corresponding to the error text and means for replacing the error text with the correct text.

2. The camera of claim 1 wherein the optics is adjustable to provide variable depth of field.

3. The camera of claim 1 wherein the text linker appends the text data tiles in proper alignment to an electronic file.

4. The camera of claim 3 further comprising

image stitching engine for generating a graphics figure in response to stitching some of the plurality of image data tiles that contain portions unrecognized by the optical character recognition engine; and wherein the linker is further configured to append the graphics figure to the electronic file.

5. The camera of claim 1 wherein the image tiles are of different spatial resolutions.

6. The camera of claim 1 wherein each image tile has the full spatial resolution of the camera imager.

7. The camera of claim 1 wherein the character recognition engine and the text linker are implemented as a processor executing instructions.

8. The camera of claim 1 further comprising

an output interface to a data processing device which is not exclusively a stand alone camera, for transferring the electronic file to the device.

9. The camera of claim 8 wherein the output interface complies with a computer peripheral bus specification.

11. An article according to claim 1 further comprising instructions that when executed cause the system to:

generate a graphics figure in response to stitching some of the plurality of image data tiles and append the graphics figure to the electronic file.

10. An article of manufacture comprising:

a machine-readable medium having instructions that, when executed by a processor cause a system to;

receive a plurality of image data tiles that represent a plurality of light images of different portions of a scene captured by an electronic camera having optics with a depth of field greater than that of a conventional scanner;

generate a plurality of text data tiles each containing text recognized in a corresponding one of the image data tiles;

search for overlapping text in the text tiles, append the text tiles in proper alignment to an electronic file, and correct an error in the electronic file by recognizing correct text in another image tile, that represents a portion of the scene previously captured by the camera, and replacing the error with the correct text.

12. A method comprising:

forming a plurality of light images of different portions of a scene on a camera imager, the imager being a part of an electronic camera that captures 3-dimensional objects;

generating a plurality of image data tiles that represent the images using the camera;

generating a plurality of text data tiles each containing text recognized in a corresponding one of the image data tiles;

searching for overlapping text in the text tiles;

wherein at least one of the text tiles contains an error in representing text in the scene, the method further comprising generating a second image data tile of the portion of the scene that includes said text; and

then generating a second text data tile based on the second image tile and that does not contain said error and then pasting the second text tile in proper alignment into an electronic file to replace the error.

13. A method according to claim 12 further comprising

generating an electronic file to which the text data tiles are appended.

14. A method according to claim 13 wherein the generating of the text data tiles is performed by a data processing device separate from the camera.

15. A method according to claim 13 further comprising

transferring the electronic file to a data processing device separate from the camera using a computer peripheral bus.

16. A method according to claim 12 further comprising

generating a graphics figure in response to stitching some of the plurality of image data tiles that contain non-text portions unrecognized by the optical character recognition engine;

and

appending the graphics figure to the electronic file.

17. A method according to claim 12 further comprising:

changing the size of the text in the second image data tile to be substantially different from the size of the text in the image data tiles.

18. A method according to claim 17 wherein the step of changing the size of the text includes changing a focal length of camera optics without substantially changing the distance between the camera imager and an object in the scene.

19. A method according to claim 17 wherein the changing of the size of the text includes changing the distance between the camera imager and an object in the scene.

20. An electronic camera comprising:

optics to form a plurality of light images of different physical portions of a scene bearing text;

camera imager to receive the light images and in response provide a plurality of image data tiles corresponding to the plurality of light images, respectively;

optical character recognition engine to provide a plurality of text data tiles each containing a portion of said text recognized from a corresponding one of the image data tiles;

text linker to search for an overlapping portion of said text in the text tiles and to append the text data tiles in proper alignment, based on the overlapping portion, to an electronic file; and

an error correction routine that when executed corrects error text, in the plurality of image data tiles, by recognizing correct text in another image data tile obtained by the imager and that represents a portion of the scene previously captured using the imager and replaces the error text with the correct text.

21. An article of manufacture comprising:

a machine-readable medium having instructions that, when executed by a processor cause a system to;

receive a plurality of image data tiles that represent a plurality of light images, respectively, of different physical portions of a scene that includes text, captured by an electronic camera having optics with a depth of field greater than that of a conventional scanner;

generate a plurality of text data tiles each containing a portion of said text recognized in a corresponding one of the image data tiles; and

search for an overlapping portion of said text in the text tiles and on that basis append the text tiles in proper alignment to an electronic file, and correct an error in the appended text tiles by recognizing correct text in another image tile that represents a portion of the scene previously captured by the camera and replacing the error with the correct text.

22. A method comprising:

forming a plurality of light images of different physical portions of a text bearing scene on a camera imager, the imager being a part of an electronic camera that captures 3-dimensional objects;

generating a plurality of image data tiles that represent the plurality of light images, respectively, using the camera;

generating a plurality of text data tiles each containing a portion of said text recognized in a corresponding one of the image data tiles;

searching for a matching portion of said text in the text tiles and generating an electronic file to which the text tiles are appended based on the matching portion;

determining error text in the plurality of text tiles;

re-capturing, using the electronic camera, a region of the scene that contains correct text, in another image data tile;

generating another text data tile that does not contain the determined error and that does contain the correct text as recognized from said another image data tile; and

then

pasting said recognized, correct text into the file.

23. The method of claim 22 wherein the determined error text has one of improperly recognized text and a misalignment of text tiles.

24. The method of claim 22 wherein the error is determined by a user comparing the file being built on a display to a document in the scene.

25. The method of claim 24 wherein the error is determined prior to all of the text in the document being processed by the camera.

26. The method of claim 22 wherein the error is determined by a spell checker built into the camera and that can detect misspelled words in the file.

27. The method of claim 22 further comprising:

providing said another text data tile for display to a user of the camera.